Google has added a Generative AI performance report to Search Console. It breaks out impressions from AI Overviews and AI Mode separately, filtered by page, country, and date. If you run AEO campaigns, this is the first first-party signal you have that connects a specific URL on your site to its appearance inside an AI-generated response.
What the Report Is and When It Launched
The Generative AI section sits inside the existing Performance report in Search Console. Google began rolling it out in mid-2025 as AI Overviews expanded globally and AI Mode moved out of limited testing. The report surfaces separately from the standard web, image, and video tabs, which makes the data easier to isolate rather than blended into general organic metrics.
The timing matters because AI Overviews and AI Mode operate differently. AI Overviews appear at the top of standard search results. AI Mode is a separate, conversational interface. Lumping them together would obscure which surface is actually citing your content.
What Data the Report Exposes
The core metrics are impressions, broken down by:
- Page: which URLs on your site appeared inside an AI-generated response
- Country: where those impressions occurred
- Date: trends over time, comparable to how you track organic impressions today
You can filter by AI Overviews or AI Mode independently. That separation is genuinely useful. A page ranking inside AI Overviews in the US may perform very differently inside AI Mode, and treating them as one number would mask that.
| Dimension | Available |
|---|---|
| Page-level impressions | Yes |
| Country breakdown | Yes |
| Date range | Yes |
| AI Overviews vs AI Mode split | Yes |
| Clicks | No |
| CTR | No |
| Query-level data | No |
How to Access the Report
Inside Search Console, go to Performance and look for the Generative AI tab at the top alongside Web, Image, and Video. If your property doesn't show it yet, the rollout is still in progress. Once visible, the interface works the same way as any other Performance filter: set a date range, apply page or country filters, and export to CSV if you want to join the data against your internal analytics.
Start by exporting the full page list sorted by impressions. That list tells you which pages Google's AI systems are already treating as authoritative enough to cite.
What's Still Missing
The gap is significant. No clicks, no CTR, and no query data.
Clicks matter because you cannot tell whether an AI Overview impression drove any traffic. CTR is impossible to calculate without both numbers. And query data would tell you which actual searches triggered your citations, which is where most content optimization decisions start.
This means the GSC Generative AI report is a confirmation tool more than a discovery tool. It tells you that a page is being cited. It does not tell you why, in response to what question, or whether anyone clicked through after seeing it.
Combining GSC Data with Third-Party AEO Tools
The query gap is where third-party AEO tools earn their keep. Tools like Otterly, PromptWatch, and OmniSEO let you track whether your brand or content appears in response to specific prompts. That prompt-level data is something GSC will not give you.
A practical pairing: use GSC Generative AI impressions to identify which pages are already being cited, then run those pages through a citation tracker to understand which questions are surfacing them. That combination moves you from "this page is cited" to "this page is cited when someone asks X."
For larger operations, Profound and Evertune AI offer share-of-voice reporting across AI surfaces that can be mapped against your GSC impression trends. If GSC shows rising impressions on a URL and Evertune shows declining share of voice for the topic that URL covers, something in the citation pattern shifted. You can investigate from there.
SerpApi's AI Overview API is worth running alongside this for query-level visibility into what triggers AI Overviews on specific searches, since GSC won't give you that dimension.
A Practical Workflow
- Export the GSC Generative AI page report weekly.
- Sort by impressions descending. Flag pages in the top quartile.
- Cross-reference against your organic traffic. Pages with high AI impressions but flat organic traffic are not converting their citations into clicks.
- Run the flagged URLs through a prompt-tracking tool to identify which questions are surfacing them.
- Review the content on those pages against what AI Mode and AI Overviews tend to cite: direct answers, cited sources, structured data, and clear entity signals.
- Pages with zero AI impressions that target informational queries are your next optimization candidates. They're not being cited yet.
This workflow takes about two hours per week once it's set up. The GSC export is the anchor. The third-party tools fill the query and brand context.
Why This Is Now a Baseline Metric
Every AEO campaign needs a way to answer: is Google's AI citing my content? Until this report launched, the answer required either third-party tools or manual prompt testing. Neither scales well across a large site.
GSC Generative AI impressions are imperfect but authoritative. They come from Google directly, they are page-specific, and they separate the two main AI surfaces. Clients and stakeholders can understand impression trends the same way they understand organic impressions today.
Add this report to your standard monthly reporting now. The pages appearing with zero AI impressions today are the ones your content strategy should prioritize next quarter.